Jiaying Shi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Jiaying Shi has a total of 14 co-authors. Here are the top co-authors by number of shared publications:
- Shmuel S. Oren
Department of Industrial Engineering and Operations Research, University of California at Berkeley, Berkeley, USA
2 shared publications - Chen Shen
Tsinghua University
1 shared publication - Chenwei Zhang
Amazon, Seattle, WA, USA
1 shared publication - Hanbo Wang
Amazon.com, Seattle, WA, USA
1 shared publication - Jim Chan
Amazon.com, Seattle, WA, USA
1 shared publication - Jin Li
Amazon, Seattle, WA, USA
1 shared publication
Latest Publications
- Open Access
- 21 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 4678-4685
Explainable and Coherent Complement Recommendation Based on Large Language Models
A complementary item is an item that pairs well with another item when consumed together. In the context of e-commerce, providing recommendations for complementary items is essential for both customers and stores. Current models for suggesting complementary items often rely heavily on user behavior data, such as co-purchase relationships. However, just because two items are frequently bought together does not necessarily mean they are truly complementary. Relying solely on co-purchase data may not align perfectly with the goal of making meaningful complementary recommendations. In this paper, we introduce the concept of "coherent complement recommendation", where "coherent" implies that recommended item pairs are compatible and relevant. Our approach builds upon complementary item pairs, with a focus on ensuring that recommended items are well used together and contextually relevant. To enhance the explainability and coherence of our complement recommendations, we fine-tune the Large Language Model (LLM) with coherent complement recommendation and explanation generation tasks since LLM has strong natural language explanation generation ability and multi-task fine-tuning enhances task understanding. Experimental results indicate that our model can provide more coherent complementary recommendations than existing state-of-the-art methods, and human evaluation validates that our approach achieves up to a 48% increase in the coherent rate of complement recommendations.
- 21 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 4678-4685
Open AccessExplainable and Coherent Complement Recommendation Based on Large Language Models
A complementary item is an item that pairs well with another item when consumed together. In the context of e-commerce, providing recommendations for complementary items is essential for both customers and stores. Current models for suggesting complementary items often rely heavily on user behavior data, such as co-purchase relationships. However, just because two items are frequently bought together does not necessarily mean they are truly complementary. Relying solely on co-purchase data may not align perfectly with the goal of making meaningful complementary recommendations. In this paper, we introduce the concept of "coherent complement recommendation", where "coherent" implies that recommended item pairs are compatible and relevant. Our approach builds upon complementary item pairs, with a focus on ensuring that recommended items are well used together and contextually relevant. To enhance the explainability and coherence of our complement recommendations, we fine-tune the Large Language Model (LLM) with coherent complement recommendation and explanation generation tasks since LLM has strong natural language explanation generation ability and multi-task fine-tuning enhances task understanding. Experimental results indicate that our model can provide more coherent complementary recommendations than existing state-of-the-art methods, and human evaluation validates that our approach achieves up to a 48% increase in the coherent rate of complement recommendations.
- 13 November 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Systems
- Vol. 33 (3) , 3315-3324
Stochastic Unit Commitment With Topology Control Recourse for Power Systems With Large-Scale Renewable Integration
In this paper, we model topology control through transmission switching as a recourse action in the day-ahead operation of power systems with large-scale renewable generation resources. We prove that transmission switching could only reduce the linear objective value of direct current optimal power flow (DCOPF) when congestion exists. However, we also show, through a simple example, that unit commitment cost could be reduced by transmission switching even in the absence of congestion. To solve the stochastic unit commitment with topology control within reasonable computational time, we proposed a heuristic that first decomposes a practical system into zones and then solves the problem for each zone in parallel. The benefit of topology control recourse is demonstrated on a network representing the Central European System. We compare the costs of the network with different loading and renewable generation conditions. The cost reduction of the test system can reach 3.34% with heavy load and large-scale renewable generation while in a single zone the cost reduction can be above 7%.
- 13 November 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Systems
- Vol. 33 (3) , 3315-3324
Stochastic Unit Commitment With Topology Control Recourse for Power Systems With Large-Scale Renewable Integration
In this paper, we model topology control through transmission switching as a recourse action in the day-ahead operation of power systems with large-scale renewable generation resources. We prove that transmission switching could only reduce the linear objective value of direct current optimal power flow (DCOPF) when congestion exists. However, we also show, through a simple example, that unit commitment cost could be reduced by transmission switching even in the absence of congestion. To solve the stochastic unit commitment with topology control within reasonable computational time, we proposed a heuristic that first decomposes a practical system into zones and then solves the problem for each zone in parallel. The benefit of topology control recourse is demonstrated on a network representing the Central European System. We compare the costs of the network with different loading and renewable generation conditions. The cost reduction of the test system can reach 3.34% with heavy load and large-scale renewable generation while in a single zone the cost reduction can be above 7%.
- 11 August 2016
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-7
Wind power integration through stochastic unit commitment with topology control recourse
We introduce a two stage stochastic unit commitment formulation in which the second stage recourse actions include possible reconfiguration of the power grid topology through transmission line switching. We conduct numerical test on IEEE 118 system to demonstrate the benefits of such adaptive topology control. Our results show that such recourse capability in response to realized uncertainty of intermittent renewable resources can mitigate such adverse variability and significantly improve unit commitment efficiency.
- 11 August 2016
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-7
Wind power integration through stochastic unit commitment with topology control recourse
We introduce a two stage stochastic unit commitment formulation in which the second stage recourse actions include possible reconfiguration of the power grid topology through transmission line switching. We conduct numerical test on IEEE 118 system to demonstrate the benefits of such adaptive topology control. Our results show that such recourse capability in response to realized uncertainty of intermittent renewable resources can mitigate such adverse variability and significantly improve unit commitment efficiency.
- 1 February 2013
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Impact of DFIG wind power on power system small signal stability
Nowadays renewable energy provides about 1/5 of the electricity generation worldwide. Among all the renewable generation forms, wind power generation is regarded as one of the most developed and efficient ways to utilize renewable energy. With more and more wind power integrated into high voltage transmission grid, whether wind power will influence the small signal stability of power systems, what kind of impact it will bring about and how wind power will influence the system small signal stability attract more and more attention. In this paper, an analytical method will be proposed to analyze the impact of DFIG integration to power system oscillation modes and mode shapes. The foundation of this method is sensitivity analysis. Simulations are conducted on a 4-machine-11-bus test system. Simulation results show that the DFIGs introduce no poorly damped electromechanical modes. The states of DFIG hardly contribute to the electromechanical modes. Furthermore, the parameters of DFIG almost have no influence on the modes.
- 1 February 2013
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Impact of DFIG wind power on power system small signal stability
Nowadays renewable energy provides about 1/5 of the electricity generation worldwide. Among all the renewable generation forms, wind power generation is regarded as one of the most developed and efficient ways to utilize renewable energy. With more and more wind power integrated into high voltage transmission grid, whether wind power will influence the small signal stability of power systems, what kind of impact it will bring about and how wind power will influence the system small signal stability attract more and more attention. In this paper, an analytical method will be proposed to analyze the impact of DFIG integration to power system oscillation modes and mode shapes. The foundation of this method is sensitivity analysis. Simulations are conducted on a 4-machine-11-bus test system. Simulation results show that the DFIGs introduce no poorly damped electromechanical modes. The states of DFIG hardly contribute to the electromechanical modes. Furthermore, the parameters of DFIG almost have no influence on the modes.